What You Do on Monday Morning

This is the final post in a six-part series on rebuilding the agency model for the autopilot era. If any of this resonated, we work with founder-led agencies at $1M-$10M on exactly this transition at Agency Focus. The conversation usually starts with one question: Which service line do you restructure first?

So here is what the first 90 days of this transition look like for an agency doing $2–5 million in revenue. Not the full transformation. Just the moves that do not break the business while you rewire it underneath.

I am being specific about the revenue range because the transition looks different at different sizes. Below a million, you do not have enough process to restructure yet. You are still finding your niche and landing clients. Above ten million, you have organizational complexity that makes the sequencing harder. The two- to five-million range is the sweet spot where you have enough clients to see patterns, enough team to delegate, and enough revenue to absorb the short-term friction of changing how you work.

If you are outside that range, the principles still apply. The timeline will be different.

You are not transforming your entire agency in ninety days. You are picking one service line and running the transition on that single slice.

The criteria for choosing are simple. It should be your most repeatable service. The one where you already know the steps because you have done it hundreds of times. The one where a new team member could shadow you for two weeks and reach 80% of your quality. The one where the outcome is measurable, and the client already cares about the number.

For a demand gen agency, this might be outbound prospecting. For a dev shop, it might be a specific type of build you do repeatedly. For a brand agency, it might be the launch playbook you run for every new client. The point is to pick the service with the highest intelligence-to-judgment ratio, because that is the service where systems can replace the most human effort the fastest.

Do not pick your most creative service. Do not pick the service you are most proud of. Pick the one that is most systematizable. Pride is not a selection criterion. Repeatability is.

In these two weeks, you should also do one other thing. Go back to Post 1 and run the honest audit on this specific service line. Map every deliverable on the intelligence-judgment spectrum. Write down, on paper, what percentage of the work is pattern recognition and what percentage is genuine strategic taste. If the intelligence percentage is below sixty, pick a different service line. You want to start where the math is most in your favor.

By the end of week two, you should have one sentence written down: "The service line we are restructuring is [X], and approximately [Y] percent of the delivery is intelligence work that a system could handle."

That sentence is your mandate for the next eleven weeks.

This is the part where most agencies lose momentum because it feels like overhead. It is not overhead. It is the foundation on which everything else sits.

Take the service line you picked and document every step of it. Use the four artifacts from Post 3.

Write the markdown files. One per task. Input, process, output, quality check. Every decision point. Every tool. Every exception. Written as if someone starts Monday and needs to deliver by Friday without asking you a single question. You will not get this right on the first pass. That is fine. Write the first version, hand it to your most junior team member, watch them try to follow it, and then rewrite the parts where they got stuck. That second version is the real one.

Record the screen. Every task that involves a cursor gets a Loom or a screen recording. Written instructions miss the spatial reasoning. Five minutes of video saves five hours of questions later.

Start the decision log. From this point forward, every judgment call on this service line gets documented. Date, decision, who made it, why, and what alternatives were considered. This feels tedious in week three. By month six, it is the most valuable document in your agency.

Start the failure file. Go back through the last twelve months of this service line and write down every engagement that underperformed. What happened? Why? What would you do differently? This retrospective is uncomfortable. It is also the fastest way to identify the patterns that your system needs to account for.

By the end of week six, you should have a complete documentation package for this one service line. Not perfect. Complete. The difference matters. Perfect documentation is a trap that lets you avoid the next step. Complete documentation is good enough to hand to someone else.

You are not repricing your entire book of business. You are restructuring the pricing on one new deal. One.

Take the next prospect who comes in for the service line you picked and pitch them the outcome model. Setup fee, monthly retainer tied to a defined metric, performance layer above target. The pricing structure from Post 2. The proposal structure from Post 5. The objection handling you rehearsed.

This is going to feel uncomfortable. You are going to want to fall back to the hourly model because it is familiar, the prospect is right there, and the deal is real. Do not fall back. Run the pitch the new way and see what happens.

The prospect says yes. You now have one client on the new model, and you can start learning what works and what breaks in real delivery against an outcome guarantee. This is the best case, and it happens more often than you think, because most prospects are tired of buying hours and are waiting for someone to offer them something better.

The prospect pushes back but stays engaged. This is almost as good as a yes because the pushback teaches you which parts of your pitch need work. The objections from Post 5 will show up here. Handle them. Adjust. The second pitch will be better than the first.

The prospect walks away. This stings, but it tells you something important. Either the prospect was a labor buyer who was never going to be your client in the new model, or your pitch needs work. Debrief honestly. Write up what happened. Add it to the failure file. Move on to the next prospect.

The goal of this phase is not to close the deal. The goal is to run the new sales process once, all the way through, and learn from it. The close is a bonus. The learning is the point.

During these same weeks, take the documentation package from the sprint and hand over the delivery of this service line to your delivery operator. Not the new deal. Your existing clients on this service line. Let someone else run the documented process while you focus on the new pitch. This is the first test of whether your documentation actually works in the real world. If the quality holds, you are ready for the next phase. If it drops, your documentation has gaps that you need to fix before you go further.

If you do not have a delivery operator, hire one. If you have someone on your team who could fill the role, promote them into it. This person takes over the service line you documented and runs it using the system you built.

The job description is simple. Follow the process. Deliver the output. Flag when something does not match the documentation. Do not improvise. Do not improve. Follow the system and tell me where the system is wrong.

That last part is the most important instruction you will give this person. You are not hiring them to do the work the way you would. You are hiring them to do the work the way the system says to, and to be honest about where the system fails. Every failure they flag is an improvement to the system. Every time they improvise without flagging it, it is a gap in your documentation that will bite you later.

This hire should be able to deliver at 80% of your quality within 2 weeks. If they cannot, go back to the documentation. The problem is almost never the person. It is almost always the process.

By the end of week thirteen, you should have one service line fully documented, one delivery operator running it, one deal pitched on the new pricing model, and a decision log and failure file that are starting to accumulate real data. Your data moat has officially started compounding.

Four numbers tell you whether the transition is working.

Delivery quality on the restructured service line, measured by client feedback and outcome metrics. If quality falls below 75% of your personal standard, the documentation needs work.

The time the founder spends on this service line, measured in hours per week. This number should go from 20+ in week 1 to under 5 by week 13. If it does not, you are still the ceiling, and something in the delivery handoff is not working.

Client retention on the restructured service line. If clients start churning at a higher rate than your baseline, the system is not delivering at the level they are used to. Investigate before you scale.

Margin on the restructured service line. If you price based on outcomes and the delivery is handled by a system with a delivery operator rather than the founder, your margin should be higher than it was under the old model. If it is not, your pricing is wrong, or your delivery cost is higher than you estimated.

Those four numbers, quality, founder time, retention, and margin, are the dashboard for the transition. Everything else is noise.

I am not going to pretend this goes smoothly. Here is what breaks and what to do about it.

Your best client on this service line will notice the change. The deliverables will feel slightly different. The communication cadence will shift. They will ask, "Is everything okay?" The answer is yes, and you should tell them what you are doing. Not the full thesis. Just "we are restructuring how we deliver this work so we can serve you more consistently." Most clients will respect the honesty. A few will be nervous. Give the nervous ones extra attention for the first month.

Your team will resist the documentation discipline. Writing down every step of a process they do on autopilot feels like busywork. They are wrong, but they are not going to believe you until they see a new person follow the documentation and deliver at quality. Once that happens, the resistance usually fades because they realize the documentation protects them, not just you.

You will want to skip the failure file. Writing down what went wrong is uncomfortable. It feels like building a record of your mistakes. It is. That record is also the reason your system gets smarter every month while your competitor's stays the same. Do not skip it.

You will want to restructure everything at once. You picked one service line for a reason. Resist the urge to expand to a second one before the first one is running without you. The transition on service line two will go twice as fast because you will know what you are doing. But only if service line one actually works first.

If you followed this sequence, you now have one service line running as a services-as-software model. Documented. Systematized. Priced on outcomes. Delivered by an operator. Measured by a dashboard. Accumulating data.

That one service line is the proof of concept for the rest of your agency.

For the next ninety days, you pick the second service line and run the same playbook. It goes faster because your team has seen the pattern. The ninety days after that, you pick the third. Within a year, your agency looks fundamentally different. Fewer people. Higher margins. Stickier clients. A data asset that compounds every month.

None of that is theoretical. I have watched agencies in this revenue range make this transition. The ones who succeed all have one thing in common: they started with one slice, they documented obsessively, and they resisted the urge to scale before the foundation was solid.

The ones who failed also share a common trait: they attempted to transform everything at once, which ultimately disrupted the client experience.

Start small. Start Monday. Start with the one service line where the intelligence-to-judgment ratio gives you the best odds. Write the first Markdown file. Record the first Loom. Start the decision log. Price the next deal based on outcomes. Hand the delivery to someone other than you.

That is the whole transition in one paragraph. The rest is repetition and patience.

This is the final post in a six-part series on rebuilding the agency model for the autopilot era. If any of this resonated, we work with founder-led agencies at $1M-$10M on exactly this transition at Agency Focus. The conversation usually starts with one question: Which service line do you restructure first?